Adding independent bias on a cathode bias amp

[Champ_49]

Hi everyone.
I am currently trying to add two potentiometers that can individually bias each el84 cathode biased amplifier.  A schematic was provided on the London power forum regarding this along with a wealth of information that I am still deciphering.

As of now I just want to get this working.

The amplifier in question is a modulus kit which is cathode biased with a master volume.
 I have attached a layout based on the schematic.

However the bias pots are not adjusting the current to the tubes with enough range or something seems off. The range from one end of the pot to the other is a difference of 0.8mV at most. 

I added test points to measure it via a 1ohm resister from the cathode pin 3.
Any idea what could be going on here?

I have also increased the value of the cathode resister Rk from 150R to 300R.

Any help would be very appreciated.

Thanks.

[kagliostro]

A schematic was provided on the London power forum .............

can you post a link ?

Franco

[jeff]

I'd suggest NOT using pin 1 as a tie point. SOME EL84's have internal connections to pin 1.
If you want individual biasing:
Separate the two tubes cathode resistor.
If the amp uses a shared 150 ohm for both tubes, use individual 300ohm for each tube.
Wire pots in series with 300ohm resistor(maybe lower 300ohm to 270ohm for adjustment both ways) to increase resistance of cathode resistor to adjust bias.
Be SURE to use pot rated to handle current, not a normal vol/tone pot, a rheostat, one rated to handle plate/screen current as all plate/screen current will go through these pots.


This is how I'd do it... But wait to see what the community has to say before attempting.
Peace

[shooter]

maybe read through this thread;
https://el34world.com/Forum/index.php?topic=30358.0

[Champ_49]

Thanks everyone for chiming in.

Edit:

I apologize. At the request of the original owner of the schematic to take it off i have. I will need to honor that request.

[Champ_49]

I'd suggest NOT using pin 1 as a tie point. SOME EL84's have internal connections to pin 1.
If you want individual biasing:
Separate the two tubes cathode resistor.
If the amp uses a shared 150 ohm for both tubes, use individual 300ohm for each tube.
Wire pots in series with 300ohm resistor(maybe lower 300ohm to 270ohm for adjustment both ways) to increase resistance of cathode resistor to adjust bias.
Be SURE to use pot rated to handle current, not a normal vol/tone pot, a rheostat, one rated to handle plate/screen current as all plate/screen current will go through these pots.


This is how I'd do it... But wait to see what the community has to say before attempting.
Peace

Thanks Jeff. I always thought pin was was no connection so the grid pin2 is connected while using pin 1 as a support for the other end of the resistor?  I have two pairs of El84's. One are JJ's and the other is Tungsol current production tubes. I will double check the data sheets to see how the pins are on them. Thanks. Didn't think if that.

[Champ_49]

maybe read through this thread;
https://el34world.com/Forum/index.php?topic=30358.0

Thanks Shooter. Will take a look at that thread.

[jeff]

Bias is the difference between the voltage on the grid and on the cathode.
The grid should be negative compared to the cathode. In fixed bias a negative voltage is applied to the grid while the cathode is at ground. In a cathode bias amp, the grid is at ground and a cathode resistor is used so the cathode is at a positive voltage. Both result in the cathode being more positive than the grid. Either the Cathode at ground and grid has a negative voltage(fixed), or the grid is at ground and the cathode is positive(cathode bias), the grid is negative in reference to the cathode.


If I'm understanding your schematic correctly, it looks like your bias pots are putting positive voltage on the grids as well as a positive voltage on the cathodes. I don't understand why you'd want to do that in a cathode biased amp.


Are you trying to use a master volume circuit designed for fixed bias in a cathode biased amp?
I'm asking because it's not mix and match.
In a cathode biased amp the MV should be referenced to ground.
In a fixed bias the MV could be tied to a bias voltage... But it'd be a Negative voltage.
I guess what I don't understand is what are you trying to accomplish?


In a cathode biased amp the grid(and MV) would be at ground potential, in other words the MV pot to ground, not a voltage, and the value of the cathode resistor sets the bias. To change the bias in a cathode biased amp you would only need to change the value of the cathode resistor. This could be done with the schematic I drew.


Trying to help, just a little confused with your schematic.

[sluckey]

It's OK to put a positive voltage on the grid ***AS LONG AS*** there is a bigger positive voltage on the cathode. Some common examples are cathode followers, cathodyne PI, Long Tail Pair PI. They all have pretty big positive grid voltage. They also have even bigger positive voltage on the cathode. In all cases, the grid is still negative with respect to the cathode. All is well.

[Champ_49]

Bias is the difference between the voltage on the grid and on the cathode.
The grid should be negative compared to the cathode. In fixed bias a negative voltage is applied to the grid while the cathode is at ground. In a cathode bias amp, the grid is at ground and a cathode resistor is used so the cathode is at a positive voltage. Both result in the cathode being more positive than the grid. Either the Cathode at ground and grid has a negative voltage(fixed), or the grid is at ground and the cathode is positive(cathode bias), the grid is negative in reference to the cathode.


If I'm understanding your schematic correctly, it looks like your bias pots are putting positive voltage on the grids as well as a positive voltage on the cathodes. I don't understand why you'd want to do that in a cathode biased amp.


Are you trying to use a master volume circuit designed for fixed bias in a cathode biased amp?
I'm asking because it's not mix and match.
In a cathode biased amp the MV should be referenced to ground.
In a fixed bias the MV could be tied to a bias voltage... But it'd be a Negative voltage.
I guess what I don't understand is what are you trying to accomplish?


In a cathode biased amp the grid(and MV) would be at ground potential, in other words the MV pot to ground, not a voltage, and the value of the cathode resistor sets the bias. To change the bias in a cathode biased amp you would only need to change the value of the cathode resistor. This could be done with the schematic I drew.


Trying to help, just a little confused with your schematic.

I am trying to make an adjustable cathode bias control for it thats all.  I can see that the MV is not mix match at the moment. Because I am also trying to reference it to ground but according to the schematic ground is also connected to the bias pot. This opens the MV up and cannot be controlled. So I clipped it also to ground but then the bias control doesn't vary when I do that. So I'm trying to make this work.

 

[Champ_49]

It's OK to put a positive voltage on the grid ***AS LONG AS*** there is a bigger positive voltage on the cathode. Some common examples are cathode followers, cathodyne PI, Long Tail Pair PI. They all have pretty big positive grid voltage. They also have even bigger positive voltage on the cathode. In all cases, the grid is still negative with respect to the cathode. All is well.

Would this require a larger value cathode  resistor?

[jeff]

Ok,
Guess I don't get how/why you're using those pots in that manner.
If all you want is bias control what's the theory behind your schematic vs. just using pots to adjust cathode resistor value?
How are those bias controls?

[PRR]

What are your actual cathode and grid voltages?

We may see what is wrong from that.

If not wrong, it may ease concerns about "positive grid". If the grid is +5, and the cathode is +20, the grid IS negative of cathode. But we can't know that until you report it.

[Champ_49]

Thanks PRR.
I will check to see what it is. So basically I should measure the voltage potential between the cathode and the grid am I correct?

[Champ_49]

OK. I measured from the cathode to the grid pin2 and I am getting appromately -14V. I have corrected some portion of this circuit by removing a wire that grounded the MV to the chassis earth ground. That was the reason why the bias wouldn't adjust accordingly.  The only issue now is I am trying to isolate the MV itself from the bias circuitry so it is functioning as it should.  Currently I have  2X150R  resistors to equal 300R for the cathode resistor and with the bias pot at "zero" it starts at a current draw of about 22mV. Adjusting it from zero and up increases it but it looks like the reference point of the MV now is shifting as well. So the volume increases as I turn the bias pot up even though the MV is physically at zero.  Kevin O'Conner at London Power has said I have to capacitively couple the MV from the bias circuitry. I think I understand it as in we are trying to prevent DC from getting into and through the MV. But I'm not sure how to do that.

[kagliostro]

Coupling caps

Franco

[PRR]

...basically I should measure the voltage potential between the cathode and the grid am I correct?

Unless otherwise requested, "measure the voltage" usually implies "from common ground".

(When I was a lad, a good bench voltmeter was grounded, so un-grounded readings blew the wall fuse.)

Obviously we can figure-out any differential voltage from a set of single-end voltages.

Though there seem to be more than a few parts here; we need to work from the same hymn-book.

[passaloutre]

Separate the MV from the bias with coupling caps (prior to the bias supply). That way the AC signal can be attenuated without affecting the DC bias.

[Champ_49]

Separate the MV from the bias with coupling caps (prior to the bias supply). That way the AC signal can be attenuated without affecting the DC bias.

Thank you. I got some help from Kevin O'Conner from London Power and that was exactly so as you mentioned. A 470uf cap goes across from the wiper of the bias pot to pot "o".

[Champ_49]

Ok,
Guess I don't get how/why you're using those pots in that manner.
If all you want is bias control what's the theory behind your schematic vs. just using pots to adjust cathode resistor value?
How are those bias controls?

I Jeff. I wanted to follow up on your question. Initially the idea of having individual adjustments was to balance the hum.  It's working as it should now in that regard.

I have it now so that the current draw is maxed out to the stock idle current with the bias pots maxed. To ensure it doesn't draw too much current.

My next question is from what I got from Jeff cathode bias is essentially adjusting current draw in reference to the cathode. Fixed bias is adjusting in reference to ground.  Would it be worth going a step further and changing the circuitry to a fixed bias to ground. Tonally would this make a difference?

[HotBluePlates]

... from what I got from Jeff cathode bias is essentially adjusting current draw in reference to the cathode. Fixed bias is adjusting in reference to ground.  ...

No, this is a wrong way to think about the bias methods.

    Fixed Bias:  Apply a fixed, negative voltage to a grid.  Grid is negative of the cathode (which is often, but not always, at 0v).

    Cathode Bias:  Use the cathode current to create a voltage-drop across a resistor to raise the cathode "above ground."  Grid is negative of the cathode (which is not at 0v).


The important bit is that cathode bias is "self-generated" and "self-adjusting": if a tube tends to draw more current, it creates a larger voltage drop, which is a greater bias voltage, which reduces tube current.

     When you plug several tubes into the cathode-biased socket, they exhibit an idle current closer to an average for the group.

     When you plug tubes into a fixed-bias socket, they exhibit greater differences of plate current.  The bias voltage does not change in reaction to the tube's current.

[jeff]

... from what I got from Jeff cathode bias is essentially adjusting current draw in reference to the cathode. Fixed bias is adjusting in reference to ground.  ...

No, this is a wrong way to think about the bias methods.

    Fixed Bias:  Apply a fixed, negative voltage to a grid.  Grid is negative of the cathode (which is often, but not always, at 0v).

    Cathode Bias:  Use the cathode current to create a voltage-drop across a resistor to raise the cathode "above ground."  Grid is negative of the cathode (which is not at 0v).

(not arguing, just trying to get an understanding)
So, what do we have here?
This is a cathode biased amp because we are using the current through the cathode resistor to get a voltage difference between grid and cathode...
But also...
This is a fixed bias amp, because we are putting a negative voltage(relative to cathode) on the grid.


What is this?


Say you had a cathode biased amp that was biased cold. How would you adjust the bias? To my understanding, you would decrease the value of the cathode resistor to get less of a voltage drop, thus changing the voltage at the cathode, thus changing the relative bias voltage difference between grid and cathode.
Would you not?
In order to adjust bias on a cathode biased amp you would adjust the value of the cathode resistor, right?
Why wouldn't take that same approach if we wanted an adjustABLE cathode biased amp?


In other words, there's two ways to change the voltage relationship between cathode and grid. One is to make the cathode resistor smaller, one is to introduce a negative voltage(or a 'less positive' voltage) on the grid.


If you had a cold biased cathode biased amp, how would you adjust it?
Smaller cathode resistor
OR
Add a voltage divider to the cathode(which is too positive) to take a small percentage of that voltage to put on the grid, so the relationship of cathode to grid is lower?


Say you've got a cathode biased amp where the cathode is 30V and the grid is at ground. If you wanted to bias it hotter, say 20V, would you actually add a voltage divider to the 30V cathode to get 10V, put the 10V on the grid, creating a 30V cathode to 10V grid for a 20V bias?
Or would you just use a smaller cathode resistor?


If you would say, smaller resistor, then why not use a varable cathode resistor as a bias adjustment?
That's the part I don't get.
Both ways would work, right?
So, what is the advantage of biasing it cold with a bigger cathode resistor, using a voltage divider to get a smaller positive voltage, taking that relatively smaller voltage and putting it on the grid.


You wouldn't do that to adjust a cathode biased amp's bias why take that approach for an adjustABLE bias control?


Justcuz a pot's cheaper than a rheostat?
Or is there something wrong with using a rheostat to create an adjustable cathode resistor?
Seems like replacing the cathode resistor with a smaller resistor in series with a rheostat would be the eaisest, most direct way, to add an adjustable bias, keep the amp cathode biased, without introducing a mix or Cathode/Fixed bias scheme.


No?

I must be missing something


Peace

[shooter]

In other words, there's two ways to change the voltage relationship between cathode and grid. One is to make the cathode resistor smaller, one is to introduce a negative voltage(or a 'less positive' voltage) on the grid.

the 1st = self bias
the 2nd = fixed bias
very, very few examples of mixing those 2 methods, in engineering we call it KISS. pick one, NOT both


 

[PRR]

> Justcuz a pot's cheaper than a rheostat?

This.

I bought some $3 10W pots a few years back. They sucked big-time. Badly made. Not even hobbyist experimenter grade. Getting known-good 10W pots was gonna cost me $45 each. The bottom fell out of that market with computer controlled everything, so the few remaining products are premium price.

[jeff]

very, very few examples of mixing those 2 methods, in engineering we call it KISS. pick one, NOT both

This is both though, no?

That's what I mean, why are we doing both?
What am I missing?

[jeff]

> Justcuz a pot's cheaper than a rheostat?

This.

I bought some $3 10W pots a few years back. They sucked big-time. Badly made. Not even hobbyist experimenter grade. Getting known-good 10W pots was gonna cost me $45 each. The bottom fell out of that market with computer controlled everything, so the few remaining products are premium price.

Ok, this
Now I get it.

[HotBluePlates]

very, very few examples of mixing those 2 methods, in engineering we call it KISS. pick one, NOT both

This is both though, no?

No, it's just adjustable cathode bias.

   The reference for the grid is being moved from "0v" to "a point more-positive than 0v" to increase tube current.
   If we yank the tube from its socket, the grid-voltage drops to 0v regardless of pot-setting.

   If we had fixed-bias, something other than tube-current would create the bias voltage.
   If we yank the tube in this scenario, the (perhaps partial) bias voltage is still present.

   FWIW, I disagree with Shooter & would have zero issues mixing bias types (cathode & fixed) if there were some advantage.



If ever confused about whether something is "fixed-bias" or not then ask, "Can I bias this tube to Class B?"

     Class B is when the tube is biased right-at cutoff.  (Really 1-6mA is more typical, because "Cutoff" is hard to do).

     It is impossible to cathode-bias a tube (where the bias voltage is derived from plate/cathode current) to cutoff.

     It is impossible to get very much beyond "Class A power, plus distortion" from cathode-biased tubes.
     Once the "other side shuts off" increasing plate/cathode current tends to create a bias voltage that constrains plate current.

     If we cathode-bias a Twin Reverb, we can only manage ~60watts clean.
     If we fixed-bias a Twin Reverb, we can soar past 100w pretty easily.  The tubes can be driven further after "the other side is cut off."

[jeff]

   If we had fixed-bias, something other than tube-current would create the bias voltage.

Oh, o.k.
Thank you thank you thank you.
I'm beginning to understand.
This begs a question though:
Isn't the voltage at the cathode more or less constant though because of the bypass cap?
As the tube draws more or less current, wouldn't that cap keep the cathode at a fixed voltage?
If the cathode is 20V at idle, doesn't the cap maintain 20V as the tube draws more/less current?
Is tapping the cathodes 20V and sending a portion to the grid any different than having a 20V bias supply if the cathode voltage remains at a constant 20V because of the cathode bypass cap?
The bypass cap... bypasses the AC(more/less current draw across the cathode resistor=AC) from the cathode's DC voltage, right?
Please excuse my ignorance, thank you for helping me understand.

[shooter]

quote from HBP in another thread;

Cathode-bias introduces local negative feedback, unless the cathode resistor is bypassed with a capacitor.

with self-biased we typically bias close to Max tube dissipation, running full throttle.  so the tube is drawing "max current" with no AC applied.
Add AC and the "swings" typically cause the tube to "run cooler"

[jeff]

Sure,
With no cap as grid swings with AC input more/less current goes through the tube and therefore through the cathode resistor thus the cathode voltage swings too as the grid voltage swings creating local neg feedback.
If we slap a cap across the cathode resistor, the cathode voltage remains constant as the grid swings.
Is that right?


The cathode voltage remains constant, so what does the grid care if it's being fed from a 20V supply that comes from a separate winding, or if it comes from a voltage divider hanging of the bypassed cathode resistor?


That's kinda what I'm getting at.


So we use a cathode resistor to bias the tube(cathode bias)
Increase the value to bias it cold.
Use a voltage divider to get smaller positive voltage relative to the cathode voltage.
Then apply that to the grid(fixed bias).


Isn't that both?
Cold cathode bias that you adjust by adding fixed bias?

[PRR]

> Isn't the voltage at the cathode more or less constant though because of the bypass cap?

Look at extremes.

If you have a big cap and a small/short transient (most music), the voltage won't change much.

If you need to drive a AC motor at a strange frequency at high power for some minutes, no cap can hold charge forever, and maybe not for long.

[HotBluePlates]

... If we slap a cap across the cathode resistor, the cathode voltage remains constant as the grid swings.
Is that right? ...

We say, "a capacitor passes AC," but really nothing passes through the capacitor (due to its dielectric).  The capacitor just appears to pass AC because electrons can be pulled onto its plates to charge the cap.

If capacitance is not infinite, then the charge (electrons) it stores will be finite.
Current is just those electrons moving, maybe out of the capacitor.

When one side of push-pull shuts off, charge is being sucked out of the cap and there is finite time before Stored Charge (voltage) falls.

Whether the cap sustains constant-voltage depends on:
   How many µFs.
   How big a current-change happens that sucks out Charge.
   How long.

... what does the grid care if it's being fed from a 20V supply that comes from a separate winding, or if it comes from a voltage divider hanging of the bypassed cathode resistor?
...
Use a voltage divider to get smaller positive voltage relative to the cathode voltage.
Then apply that to the grid(fixed bias).

No, because:

   If we yank the tube from its socket, the grid-voltage drops to 0v regardless of pot-setting.

And that is because:

... cathode-bias a tube (where the bias voltage is derived from plate/cathode current) ...

No plate current, no bias voltage = "cathode bias"
No matter how an individual perceives the "adjustable cathode bias setup," it will not work deep into Class AB for the reasons I outlined earlier.

"Applying stuff to the grid" is not a definition of any bias-type.

[jeff]

Whether the cap sustains constant-voltage depends on:
   How many µFs.
   How big a current-change happens that sucks out Charge.
   How long.

You're talking frequency of the note, not duration, right?
If the cap is big enough for 82Hz and you hit an E, it'll react the same in a short pluck as it would for a long sustained note.
Is that what you mean?


If your cap is big enough to handle your instrument's lowest frequency, will your cathode will always be at a constant voltage?


If we use a smaller cap, we introduce nfb frequencies below the caps range will cause cathode voltage to vary.


Is this right?
Not how long the note is played, but the wave length... the frequency?

[passaloutre]

I’m pretty sure it means the length of the note. A capacitor is like a battery, it can only hold so much charge.

[pdf64]

…
If your cap is big enough to handle your instrument's lowest frequency, will your cathode will always be at a constant voltage?
…

Yes, signal frequency, not duration.
With push pull output stages, cathode voltage will increase when the signal level pushes operation into class B / overdrive. Due to the fall in OT primary impedance when one half primary or the other is in cut off. 

[HotBluePlates]

Whether the cap sustains constant-voltage depends on:
   How many µFs.
   How big a current-change happens that sucks out Charge.
   How long.

You're talking frequency of the note, not duration, right?

Both matter.

Frequency & circuit-resistance affects charge/discharge time.
Push-pull Class A is small-current (maybe 20mA peak for 6V6?), small-duration of overload, short-time the cap has to supply current.
Push-pull Class AB with big-tube (maybe 200mA peak for 6L6?), long-duration of note, long-time the cap has to pull big charging current through circuit-resistance.


Review Capacitor Charge/Discharge, Time Constant, etc.  Without realizing it, you're mis-applying assumptions you developed based on how a cathode-biased preamp tube behaves.  Power tubes are bigger current, often pushed harder/further than preamp tubes; suddenly it's not just the time-constant that matters.

Even if you don't fully understand, tack a voltmeter across a power section cathode bypass cap & blast that amp as loud as you can.  The "constant cap voltage" will fall before your eyes.  If you make the cathode resistor 3-4 times bigger (to simulate trying to bias into Class AB), the cap voltage will fall even faster (because the tube is trying to pull more instantaneous current from the cap).

[Champ_49]

From my understanding this is an adjustable cathode biased amp. I think you'd call it push pull? As I adjust one pot it increases or decreases current through the tubes. But decreasing or increasing one side would affect the other tube.

The grid has zero volts bit it's negative with respect to the cathode by the resistance value. Hence cathode biased.

If I'm understanding this correctly (I'm still a novice and learning). In a fixed biased amp the cathode is grounded and the grid is negative via negative voltage applied through the grid leaks? Vs a cathode bias where grid is negative but in relation to the cathode.

[jeff]

This is fascinating stuff.
Can I ask a(nother) dumb question?
Are these/could these panel mount controls?
Could you tune the range of the pot so you could go from coldest limit to hottest limit, and would this be a sort of tone knob? Clean power/earlier breakup?

[Champ_49]

I'm new to all this as well. But I have it set up with panel mount pots. I'm not sure what others will think but I am using 1/10W rated pots so save space and make it inconspicuous. 

I calculated according to the formula provided.

P=V^2/R

V is about 43V across the pot.
R I'm using 50K pot

So P=43X43/50K = 0.03698W = 36.98mW

This pots rated 0.1W = 100mW

This should be more than sufficient to handle the power in this circuit?  Even with the power times 2 = 73.96mW for safety factor.

[shooter]

when in doubt, keep a fire extinguisher close, no cigars, they mask the smell of fried electronics 


is the pot carbon coated plastic?  you might get warpage from heat over time

[HotBluePlates]

Are these/could these panel mount controls?
Could you tune the range of the pot so you could go from coldest limit to hottest limit, and would this be a sort of tone knob?

Panel mount, Yes.
"Tone knob," No.


In my case, I used a version of this "balance-able" cathode bias circuit, with a 2w pot wired as a variable resistor between the cathode & the 68Ω resistor.  That pot adjusted overall tube-current, the other pot (labeled "100 ohm 3w/ww" in the diagram) adjusted the balance of idle current between the tubes (same as what this thread has discussed, but implemented differently).

I used panel-mount locking pots, which are positioned between the 2 output tubes in the photo below.  Back off the lock-nut slightly, and they're screwdriver-adjust.  The Black & Red pin-jacks are for monitoring individual tube current for the adjustment.

[Champ_49]

when in doubt, keep a fire extinguisher close, no cigars, they mask the smell of fried electronics 


is the pot carbon coated plastic?  you might get warpage from heat over time

Hey shooter. Not sure. I. Confused because on the mouser it says "element: metal".

Never heard of metal as an element for pots. But I heard of cermet.

[Champ_49]

Are these/could these panel mount controls?
Could you tune the range of the pot so you could go from coldest limit to hottest limit, and would this be a sort of tone knob?

Panel mount, Yes.
"Tone knob," No.


In my case, I used a version of this "balance-able" cathode bias circuit, with a 2w pot wired as a variable resistor between the cathode & the 68Ω resistor.  That pot adjusted overall tube-current, the other pot (labeled "100 ohm 3w/ww" in the diagram) adjusted the balance of idle current between the tubes (same as what this thread has discussed, but implemented differently).

I used panel-mount locking pots, which are positioned between the 2 output tubes in the photo below.  Back off the lock-nut slightly, and they're screwdriver-adjust.  The Black & Red pin-jacks are for monitoring individual tube current for the adjustment.

I have it similar to yours. The current measurement points are in the rear of the chassis. As of now I have it wired up on a breadboard. But by the end of tomorrow I should have it all wired up inside the chassis.

[Champ_49]

This is fascinating stuff.
Can I ask a(nother) dumb question?
Are these/could these panel mount controls?
Could you tune the range of the pot so you could go from coldest limit to hottest limit, and would this be a sort of tone knob? Clean power/earlier breakup?

There are two Rks in series in my setup. I kept the RK1 stock at 150R and the other at around 680R. With that setup the bias range is 10mV to 34mV per tube. I can adjust it for colder by increasing Rk2 and also even hotter by decreasing Rk1. But the voltage across the two resistors also increases and can increase significantly so that would mean bigger sized caps which i dojt have the space for now. I personally like this range.